Longitudinal PET: Lacunes, Cognition and Behavior This project uses longitudinal PET to address important questions about diagnosis, prognosis, and the mechanisms of cognitive failure in these patients. The grant uses a longitudinal design and well established methods to address question that have yielded little progress despite decades of scientific attention. For example, it is presently extremely difficult to determine the relative contributions of cerebrovascular pathology and Alzheimer's disease in patients with cognitive dysfunction and small subcortical strokes (lacunes). Our preliminary data suggest that PET may be useful in defining prognosis and we will explore this application of PET more definitively. These clinical uncertainties reflect underlying confusion about the mechanisms through which lacunes may cause cognitive failure. The belief that lacunes that occur in brain structures critical to cognitive function ("strategic lacunes") are the primary cause of IVD is widely accepted, yet it has never been adequately tested. To do is a major goal of this grant. The addition of longitudinal LET and MRI studies to out protocol offers the opportunity to test all of these questions in unique ways that allow stronger attributions of causality. This is a prospective longitudinal study with an initial sample of 125 subjects. There will be 75 subjects with lacunar strokes and different degrees of cognitive dysfunction (normal, mild cognitive impairment, dementia)> Subjects are studied with positron emission tomography (PET) to measure regional brain metabolism, with MRI to measure brain structures, and with neuropsychological tests and behavioral scales to measure symptoms. All subjects will receive annual clinical follow-up evaluations consisting of interim medical history, neurological exams, psychiatric evaluations, and neuropsychological testing. Brain imaging studies will be repeated two years after the baseline study. This project is expected to yield better understanding of the mechanisms of cognitive failure associated with subcortical stroke, to improve clinical diagnosis, and to improve the prediction of the course of cognitive function in this population.